- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I setup a basic PRiSM block connected to the onboard LED on the development kit. I set my target Fout to 500 Hz and I see a lot of flicker. The app note about PRiSM states that "FOUT ~= 300 Hz to have flicker free LED dimming" I can't seem to get it flicker free be 2 kHz. I have validated Fout with my oscope and it does show 500 Hz.
My setting are as follows:
Fin = 2 kHz
Pulse Density = 50%
Resolution = 8 bit
Fin calculation
Fout = Fin / 4 @ 50% pulse density
500 Hz = 2 kHz / 4
- Labels:
-
PSoC 5LP
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
If you go to the start page of Creator there is an example project
that you can run. Without seeing your project archive I would suggest
this.
Regards, Dana.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
If I look at periods seems to run at 8 mS or less for any period, or
> 100 Hz. So flicker would seem it should be be OK from this.
I noticed you did not terminate reset pins with a logic "0".....?
Also you are feeding PRIsM with ~ 100 Khz clock, is that what
you intended ?
Maybe someone else can do a setup, I am on vaca and no board
available.
Regards, Dana.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
From your last posted example:
your input frequency is 2kHz, divided by 256 (you selected 8 bit resolution) results in 8Hz which will flicker. So I'd suggest you to use 24kHz input frequency which should give you a (nearly) flicker-free ~90Hz.
Bob
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
You should terminate the PriSM reset input with a logic low, dont
leave it floating.
Regards, Dana.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Ok, setting it to zero doesnt seem to rectify the flickering though.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
You changed the clock per Bob's instructions in prior post ?
Regards, Dana.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I have been trying various freqencies for a few weeks now.
Per this app note http://www.cypress.com/?docID=50255
I am at 50% duty cycle which is Fin / 4. The equation Bob suggests is for a duty cycle 0% and 100%. A 24kHz clock will result in an Fout being 24 kHz / 4 = 6 kHz. This will definitely be flicker free but way to high for my inductors.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Are you meeting these conditions -
Also makes me wonder if there is any possibility you get "beat" notes
between LEDs....Can you run just one LED to confirm this is not happening.
Regards, Dana.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
yes, Fout is 500 Hz. The latest project i have attached only contains 1 LED on the development kit board which is just an LED and resistor in series.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Try to attach project again.
Regards, Dana.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Please find attached project, project configuration below
- Single PrISM Block
- Fin = 2kHz
- Pulse Density = 50%
- Resolution 8 bit
- Fout = 500 Hz
- LED output P6[3], onboard LED on development kit
The flicker seems to get worse lower towards Fout of 300 Hz, it goes away at high frequencies but theres soo much inductor noise. I can provide more logic captures and I am able to video flicker if that helps.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I am going to advise you file a CASE and point to this thread.
Unless someone else can fire up a board and look at your
project, I can't because I am on vacation.
To create a technical or issue case at Cypress -
“Support”
“Technical Support”
“Create a Case”
You have to be registered on Cypress web site first.
Your clock, at 2 Khz, means you will have (from datasheet) -
Maximum output frequency components present for any
signal density in PrISM sequence are FIN/2 and minimum
are FIN/2n.
The average output frequency for 50% signal density is
average of FIN/2 (maximum value) and FIN/2n (minimum
value). For high n, the value of FIN/2n is low compared to
FIN/2 and average output frequency is nearly FIN/4.
So you will have worst case 2 Khz/256 or ~ 8 Hz, way into flicker.
Maybe its time to consider PWM......
Regards, Dana.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I hear what you are saying about the bounderies being minimum: 8 hz and maximum 1 kHz. The datasheet says to keep these bounderies of Fout <=1kHz but above 120 Hz. What settings would you recommend to achieve that? It seems impossible to me. I will open a case to understand whats going on.
I was interpreting that Fout was the average and not the extremes.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Why do you have an inductor in your circuit? The whole point of driving a LED with PWM or PRiSM is _not_ to use an inductor, but really run the LED pulse-based. That ensure only brighness changes, but not color / color temperatur. When you remove the inductor you can run the whole system with higher frequencies.
(The way how PRiSM works will not give you a narrow frequency range, for a 8-bit-PRiSM the span will always be 1:128)
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I am driving very high powered LEDs which cannot be driven right off the board. I am experiencing flicker on the driving of those LEDS as well as the LED on the development kit. I am hoping if I can solve this problem on the development kit, the solution will carry over to the driver board.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Since the objective of PrISM is to spread the frequency over a range as large as possible I would suggest you to use PWM.
Choose a fixed period that already guarantees flicker-free (100-200Hz) and only vary the compare-value between 0 to 100% of the period value.
Bob
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
But even with the inductor you do not drive the LEDs directly from PSoC. So you just need a FET turning the LEDs on and off, and still don't need an inductor. Or do you use an existing driver that has a PWM input?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I hear what you are saying about the bounderies being minimum: 8 hz and maximum 1 kHz. The datasheet says to keep these bounderies of Fout <=1kHz but above 120 Hz. What settings would you recommend to achieve that? It seems impossible to me. I will open a case to understand whats going on.
I was interpreting that Fout was the average and not the extremes.
The graph shows you what average F to expect. So if you are varying over
a wide range you will get clk/2**n or some pretty low avg F.
If using 8 bits its a counter that has a base period of 256 counts, So for 120 Hz
I would suggest 30 Khz. If thats a problem for the LED driver then a PWM solution
would allow you better optimization, at the expense of additional EMI.
I am a bit un unfamiliar ground myself, dying to get home to look at this myself.